Characteristics of included studies
| Study description | Study design | Patient diagnosis | Proportion of patients preferring home death (%) | Cancer cohort congruence by location (where available), overall congruence* and overall incongruence† (%) | Non-cancer cohort congruence by location (where available), overall congruence* and overall incongruence† (%) | Grading of study |
|---|---|---|---|---|---|---|
| Abarshi et al (2011)8 2008, the Netherlands Dutch sentinel network of general practitioners, n=252 | 1-year follow-back study via a general practitioner surveillance network in the Netherlands | 38% cancer | N/A | Overall: 32/72 (44.4%)‡ Incongruence: 40/72 (55.6%) | Overall: 18/93 (19.4%)‡ Incongruence: 75/93 (80.6%) | B, 65.5% preferred place of death response rate, retrospective |
| Alonso-Babarro et al (2011)9 2004 to 2006, Spain (Madrid) region health areas palliative home care teams, n=380 | 3-year prospective observational cohort study | 100% cancer | 79.8 | Home: 161/182 (88%) Other: 42/46 (91%) Overall: 203/228 (89.0%) Incongruence: 25/228 (11.0%) | B, 60.0% preferred place of death response rate | |
| Bakitas et al (2008)10 1999 to 2001, USA outpatient palliative care, n=268 | Family proxy respondents of deceased patients with breast, gastrointestinal or lung cancer who had participated in Project ENABLE§ | 100% cancer | N/A | Overall: 83/125 (66.4%)‡ Incongruence: 42/125 (33.6%) | C, 46.6% preferred place of death response rate | |
| Beccaro et al (2006)11 2000 to 2002, Italy population survey of patients who died of cancer, n=1271 | Retrospective national mortality follow-back quantitative caregiver survey (ISDOC)¶ | 100% cancer | 93.5 | Home: 614/938 (65%) Hospital: 45/49 (92%) Hospice: 1/2 (50%) NH: 13/14 (93%) Overall: 673/1003 (67.1%) Incongruence: 330/1003 (32.9%) | B, 78.9% preferred place of death response rate, retrospective | |
| Brogaard et al (2012)12 2008 to 2009, Denmark (former Aarhus County), n=96 | Prospective serial interview and questionnaire study from three specialist palliative care teams | 100% cancer | 68.8 | Home: 16/33 (48%) Hospital: 0/2 (0%) Hospice: 5/12 (42%) NH: 0/1 (0%) Overall: 21/48 (43.8%) Incongruence: 27/48 (56.2%) | C, 50% preferred place of death response rate | |
| Capel et al (2012)13 2009 to 2010, Wales regional (Cardiff) specialist community palliative care service, n=788 | Prospective observation with patient interview to determine preferences | 93% cancer | 67.3 | Home: 236/348 (68%) Hospital: 10/12 (83%) Hospice: 90/108 (83%) NH: 27/29 (93%) Other: 12/20 (60%) Overall: 375/517 (72.5%) Incongruence: 142/517 (27.5%) | Home: 27/32 (84%) Hospital: 2/2 (100%) Hospice: 3/6 (50%) NH: 6/7 (86%) Other: 1/1 (100%) Overall: 39/48 (81.2%) Incongruence: 9/48 (18.8%) | B, 71.7% preferred place of death response rate |
| Carroll (1998)14 1995 to 1996, Scotland Aberdeen, n=34 | 1-year prospective audit with general practitioner, community nurse, relatives and/or carer interview | 100% cancer | 76.7 | Home: 16/23 (70%) Hospital: 0/0 (0%) Hospice: 1/1 (100%) NH: 6/6 (100%) Overall: 23/30 (76.7%) Incongruence: 7/30 (23.3%) | B, 88.2% preferred place of death response rate, single service | |
| Dunlop et al (1989)15 1986 to 1987, England hospital palliative care support team, n=160 | Prospective observation with patient interview to determine preferences | 97% cancer | 53.3 | Home: 38/48 (79%) Hospital: 11/11 (100%) Hospice: 13/26 (50%) NH: 0/0 (0%) Overall: 62/85 (72.9%) Incongruence: 23/85 (27.1%) | Home: 0/0 (0%) Hospital: 2/2 (100%) Hospice: 0/0 (0%) NH: 0/3 (0%) Overall: 2/5 (40.0%) Incongruence: 3/5 (60.0%) | C, 56.3% preferred place of death response rate, single service |
| Fukui et al (2003)16 2001, Japan 259 home care agencies, n=528 | Nationwide study, questionnaire survey response from home care agencies selected from a list of the National Association for Home-visit Nursing Care (NAHNC) | 100% cancer | 79.4 | For spontaneously expressed home preference only: Overall: 288/419 (68.7%)** Incongruence: 131/419 (31.3%) | C, preference not assessed consistently, congruence for all locations of death not assessed | |
| Guthrie et al (1996)17 1994, England St Stephen's clinic London, n=124 | 6-month retrospective key worker interview | 0% cancer 100% HIV | 30.6 | Home: 35/38 (92%) Hospital: 15/58 (26%) Hospice: 22/25 (88%) Other: 1/3 (33%) Overall: 73/124 (58.9%) Incongruence: 51/124 (41.1%) | B, 100% preferred place of death response rate, retrospective, single service | |
| Gyllenhammar et al (2003)18 1999, Sweden palliative home care teams, n=221 | Prospective interview / questionnaire for caregivers in five palliative home care teams | 100% cancer | 37.1 | For spontaneously expressed home preference only: Overall: 63/82 (76.8%)** Incongruence: 19/82 (23.2%) | C, preference not assessed consistently, congruence for all locations of death not assessed | |
| Hsieh et al (2007)19 2007, Taiwan palliative care hospital unit, n=46 | Prospective cohort survey of patient and family, follow-up bereavement interview | 100% cancer | 73.9 | Overall: 20/46 (43.5%)‡ Incongruence: 26/46 (56.5%) | B, 100% preferred place of death response rate, single service | |
| Jeurkar et al (2012)20 2008 to 2011, USA (three hospice programs in Florida, Pennsylvania, and Wisconsin), n=7391 | 3-year electronic health record-based retrospective cohort study | 100% cancer | 88.5 | Home: 1877/3153 (60%) Other: 266/408 (65%) Overall: 2143/3561 (60.2%) Incongruence: 1418/3561 (39.8%) | C, 42.7% preferred place of death response rate, retrospective | |
| Johnson et al (2009)21 2006 to 2007, England (Bradford), n=86 | 1-year prospective study by two integrated cardiology palliative care teams, patient interview (one care team data excluded)†† | 0% cancer 100% heart failure | N/A | Overall: 22/34 (64.7%)‡ Incongruence: 12/34 (35.3%) | C, 39.5% preferred place of death response rate | |
| Johnson et al (2012)22 2009 to 2010, England (Scarborough and Bradford/Airedale), n=126 | 1-year prospective study by two integrated cardiology palliative care teams, patient interview | 0% cancer 100% heart failure | 81.2 | Home: 35/69 (51%) Hospital: 4/4 (100%) Hospice: 12/12 (100%) Overall: 51/85 (60.0%) Incongruence: 34/85 (40.0%) | B, 67.4% preferred place of death response rate | |
| Karlsen et al (1998)23 1995 to 1996, England cancer death certificates, n=229 | Random sample of cancer deaths | 100% cancer | 73.6 | Home: 37/64 (58%) Other: 18/23 (78%) Overall: 55/87 (63.2%) Incongruence: 32/87 (36.8%) | C, 38.0% preferred place of death response rate | |
| McKeown et al (2008)24 2004 to 2006, England (Liverpool), n=29 | 2-year retrospective audit assessing current referral practice for patients with end-stage renal failure | 0% cancer 100% renal failure | 50.0 | Home: 3/7 (43%) Hospital: 2/2 (100%) Hospice: 0/4 (0%) NH: 0/1 (0%) Overall: 5/14 (35.7%) Incongruence: 9/14 (64.3%) | C, 48.3% preferred place of death response rate | |
| Meeussen et al (2011)25 2008, Belgium and the Netherlands nationwide sentinel networks of general practitioners, n=422 | 1-year nationally representative retrospective follow-back study | 100% cancer | N/A | Overall: 167/219 (76.3%)‡ Incongruence: 52/219 (23.7%) | C, 51.9% preferred place of death response rate, retrospective | |
| Scholes and Allan (2012)26 2012, England (Mount Vernon Cancer Network), n=708 | 3-month network-wide audit of consecutive patients known to specialist palliative care teams | 100% cancer | 43.2 | Overall: 349/426 (81.9%)‡ Incongruence: 77/426 (18.1%) | B, 60.2% preferred place of death response rate, retrospective, single service | |
| Shega et al (2008)27 2000 to 2001, USA Geriatric clinic, n=135 | 2-year prospective study with community-dwelling dementia patient caregiver interviews, University of Chicago geriatrics clinics | 0% cancer 100% dementia | N/A | Overall: 69/104 (66.3%)‡ Incongruence: 35/104 (33.7%) | B, 77.0% preferred place of death response rate, single service | |
| Tang and McCorkle (2003)28 2001 to 2002, USA Connecticut, six study sites, terminal cancer patients, n=127 | 1-year prospective observation with patient interview | 100% cancer | 87.4 | Home: 25/111 (23%) Hospital: 3/3 (100%) Hospice: 9/10 (90%) NH: 1/3 (33%) Overall: 38/127 (29.9%) Incongruence: 89/127 (70.1%) | A, 100% preferred place of death response rate | |
| Thomas et al (2004)29 2000 to 2002, England terminal cancer patients, n=41 | 2-year prospective longitudinal observation with serial cancer patient interviews | 100% cancer | 29.4 | Home: 6/10 (60%) Hospice: 4/4 (100%) Overall: 10/14 (71.4%) Incongruence: 4/14 (28.6%) | A, 82.9% preferred place of death response rate | |
| Tiernan et al (2002)30 2002, Ireland home care team, n=191 | Prospective observational cohort study | 100% cancer | 81.8 | Home: 57/121 (47%) Hospital: 2/5 (40%) Hospice: 10/17 (59%) NH: 4/5 (80%) Overall: 73/148 (49.3%) Incongruence: 75/148 (50.7%) | B, 77.5% preferred place of death response rate, single service | |
| Townsend et al (1990)31 1986 to 1987, England hospital/community terminal cancer patients, n=70 | 1-year prospective random-sample longitudinal observation, serial patient interviews | 100% cancer | 30.5 | Home: 17/18 (94%) Hospital: 12/32 (38%) Hospice: 7/9 (78%) Overall: 36/59 (61.0%) Incongruence: 23/59 (39.0%) | A, 84.3% preferred place of death response rate | |
| Van der Heide et al (2007)32 2006, the Netherlands, n=63 | Retrospective sub-sample of a cohort of advanced cancer patients, interview with bereaved relatives | 100% cancer | 78.1 | Home: 20/25 (80%) Hospital: 2/2 (100%) Hospice: 3/3 (100%) Other: 0/2 (0%) Overall: 25/32 (78.1%) Incongruence: 7/32 (21.9%) | C, 50.8% preferred place of death response rate | |
| Walker et al (2010)33 2007 to 2010, England regional (South Devon) specialist community palliative care service, n=41 | Prospective, longitudinal observation with serial patient/carer interviews | 88% cancer | 53.7 | Overall: 17/36 (47.2%)‡ Incongruence: 19/36 (52.8%) | Overall: 1/5 (20.0%)‡ Incongruence: 4/5 (80.0%) | A, 100% preferred place of death response rate |
*Overall congruence was calculated, using data provided in the studies, by dividing the number of subjects with met preferences for any location of death by the number of subjects with a preferred location of death. This overall congruence differs from some studies’ reported congruence value because these studies reported met preferences for home death divided either by the number of home deaths or the number of preferences for home death.
†Incongruence was calculated, using data provided in the studies, by dividing the number of subjects with unmet preferences for any location of death by the number of subjects with a preferred location of death. As with reports of overall congruence, the provided values for incongruence differ from the reported values in some studies due to the variable methods of quantification.
‡Data not provided to calculate congruence individually for all locations of death: only overall congruence information available.
§Project ENABLE (Educate, Nurture, Advise, Before Life Ends): an outpatient palliative care early intervention demonstration quality improvement project implemented at two oncology care settings.
¶ISDOC study (Italian Survey of the Dying of Cancer): a national mortality follow-back survey of 2000 cancer deaths.
**Percentage of patients who spontaneously expressed a wish to die at home. No other elicitation of preference was reported. Data not provided to calculate overall congruence for all locations of death.
††The Scarborough/Whitby/Ryedale (SWR) portion of this dataset was excluded from our analyses. Preferred places of death were identified from HFNS (heart failure nurse specialist) recall, not documentation, potentially introducing recall bias.